Psilocybin for Depression: A Closer Look at the Evidence

With the rise of companies exploring psilocybin based drugs to treat various mental health conditions, the need for greater evidence based research is paramount. A new meta-analysis published in Psychopharmacology (2025) reviews the efficacy of psilocybin-assisted therapy for depression, drawing from nine randomized controlled trials (RCTs) and over 600 participants. The analysis reveals promising antidepressant effects but also exposes significant concerns around study design, risk of bias, and insufficient reporting of adverse events.

Companies developing psychedelic drugs and their stage of development. Image credits: Psychedelic Alpha

Mixed Evidence for Clinical Impact

The review found that psilocybin treatment was moderately more effective than control conditions in reducing depressive symptoms, with an effect size of g = 0.69 (95% CI: 0.34 to 1.04). This means that, on average, participants receiving psilocybin reported greater reductions in depression scores than those in control groups. However, the effect size was largely driven by small studies that used weak control comparisons—such as inactive placebos (e.g. niacin) or waitlist conditions—which are less rigorous than trials comparing psilocybin with standard antidepressants. In trials using active comparators like escitalopram, no significant difference was observed.

Heterogeneity was high across trials, meaning that the effects varied widely between studies. This could reflect differences in dosing, psychotherapy protocols, or participant characteristics. To account for possible overestimation of effect due to publication bias (i.e. the selective publication of studies with positive results), the authors applied a trim-and-fill adjustment, which reduced the pooled effect size to g = 0.56. This more conservative estimate suggests a smaller, though still meaningful, benefit.

Methodological Shortcomings

The review highlighted widespread issues with risk of bias. All included studies were rated as having either high or moderate risk across domains such as outcome measurement, allocation concealment, and selective reporting. Blinding was a particular challenge. Because psilocybin induces noticeable psychoactive effects, participants—and often therapists—could usually guess whether the active drug or placebo was administered. This expectancy bias may have inflated reported benefits.

Adverse event reporting was also inconsistent. Just two of nine studies met high-quality standards for harms assessment. Many failed to clearly define what counted as an adverse event or did not report adverse events separately for psilocybin and control groups. For example, some studies reported only mild side effects (like headaches or nausea) without noting whether these occurred more frequently in the treatment or control condition. Others lacked pre-defined monitoring protocols or criteria for halting the trial in case of serious events.

Unclear Mechanisms of Action

Although many trials speculated about the role of 5-HT2A receptor agonism, increased neural plasticity, or spiritual experiences in driving therapeutic effects, most did not measure these mechanisms directly. Mechanistic discussions were often confined to the discussion sections, rather than being part of the study's core hypotheses or design.

One exception was a follow-up neuroimaging study linked to the Carhart-Harris et al. (2021) trial, which used functional MRI to examine changes in brain network connectivity. The authors reported increased "brain network flexibility" after psilocybin treatment, a potential biomarker for antidepressant response. However, the sample size was small and the finding has not yet been replicated, limiting confidence in this proposed mechanism.

What Comes Next?

While psilocybin therapy continues to generate enthusiasm, this meta-analysis underscores the importance of rigor. Larger, independently funded trials with active controls—such as standard antidepressants or psychotherapy—and well-defined mechanistic targets are needed. These could include neuroimaging biomarkers, inflammatory markers, or psychological mediators like emotional processing or rumination.

Companies developing psilocybin based drugs. Image credits: Psychedelic Alpha

Future studies should also improve harm reporting by adopting pre-registered safety monitoring plans and standard definitions for adverse events. Better blinding techniques—such as using low-dose psychedelics or active placebos that mimic some of the subjective effects—may help reduce expectancy bias. In addition, incorporating objective outcome measures like neuroimaging or digital phenotyping could help validate self-reported improvements.

Without these steps, the field risks overinterpreting early findings from underpowered or methodologically limited studies—ultimately slowing progress toward evidence-based psychedelic treatments.

---

Citation

Borgogna NC, Owen T, Petrovitch D, Vaughn J, Johnson DAL, Pagano LA Jr, Aita SL, Hill BD. (2025). Incremental efficacy systematic review and meta-analysis of psilocybin-for-depression RCTs. Psychopharmacology. https://doi.org/10.1007/s00213-025-06788-w